Hi Doug. Using version 4 appears to have done the trick. I’ve spot-checked several clusters that were off when I was using version 3.0.3, and in each case they seem okay with 4. The identity of the clusters and the # of voxels passing my threshold doesn’t change b/w the two versions, but 4 gets the peak negative voxels correct and 3.0.3 sometimes did not.

Thanks,

Dan Dillon

Affective Neuroscience Lab

Harvard University

PS. Out of curiosity, what was the nature of the bug that was causing the problem?

-----Original Message-----
From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu]
Sent: Monday, September 24, 2007 7:01 PM
To: Dan Dillon
Cc: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] Problem concerning maximum values from mri_volcluster

Can you try it with 4? I fixed a bug that occured when reporting a negative cluster.

doug

Dan Dillon wrote:

Hi Doug. I’m working with stable version 3.0.3

-----Original Message-----
From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Doug Greve
Sent: Monday, September 24, 2007 6:26 PM
To: Dan Dillon
Cc: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] Problem concerning maximum values from mri_volcluster

Hi Dan, what version of FreeSurfer are you using?

Dan Dillon wrote:

Dear FreeSurfers,

I’m getting odd results from mri_volcluster—the max voxel listed for a given cluster often does not appear to actually be the max in that cluster. This most often appears to happen with small clusters and negative significance values.

Below is an example made with abs threshold = 3 and minsizevox = 12 (note that I’m plotting the output from mri_volcluster here, not the raw data, so all the depicted voxels have passed the threshold/cluster requirement). The cursor is at the max voxel as listed by mri_volcluster (significance value is -3.31). However, as can be seen in the example, there is a more extreme/significant value just two voxels over (laterally)—the value there is -4.03.

Below is a second example from the same analysis. In this case, I have the cursor on a voxel whose significance value is -5.39. The problem is that mri_volcluster does not list a max voxel with such a significant value—the largest (negative) significance value listed is -4.09.

When I first saw things like this I was looking at the raw data, and assumed I was seeing highly significant voxels that were not part of a large enough cluster to meet my minimum voxel size requirement. But the plots shown here were outputted by mri_volcluster using those thresholds (12 voxels, thmin = 3), so that can’t be the problem.

Am I doing something incorrect here and/or mis-reading the results of mri_volcluster? Any thoughts/advice would be appreciated. I have seen a few cases of this, not just the ones I’m showing here. Again, it seems to come up most frequently for smaller clusters and with negative significance values, although that just may be when it’s most obvious to me.

Here’s the command I used to generate the output for the examples:

mri_volcluster --in FPP_HiRisk_MC_375/bold/FPP_MCReg_8_23/tal-rfx/Rcue_v_RSF/sig --in_type bfloat --reg FPP_HiRisk_MC_375/bold/FPP_MCReg_8_23/tal-rfx/register.dat --thmin 3 --minsizevox 12 --out Test_FPPHiRisk_MC375_Rc_v_RSF --out_type bfloat

Thanks in advance.

Dan Dillon

Affective Neuroscience Lab

Harvard University

_______________________________________________

Freesurfer mailing list

Freesurfer@nmr.mgh.harvard.edu

https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

--

Douglas N. Greve, Ph.D.

MGH-NMR Center

greve@nmr.mgh.harvard.edu

Phone Number: 617-724-2358

Fax: 617-726-7422

In order to help us help you, please follow the steps in:

surfer.nmr.mgh.harvard.edu/fswiki/BugReporting

--

Douglas N. Greve, Ph.D.

MGH-NMR Center

greve@nmr.mgh.harvard.edu

Phone Number: 617-724-2358

Fax: 617-726-7422

In order to help us help you, please follow the steps in:

surfer.nmr.mgh.harvard.edu/fswiki/BugReporting